Wide Variations in Use of Testosterone Therapy Among VHA Facilities | Page 2
Published Online: October 02, 2013
Guneet K. Jasuja, PhD; Shalender Bhasin, MD; Joel I. Reisman, AB; and Adam J. Rose, MD, MSc
To test the hypothesis that there was significant variation among sites in testosterone prescribing, a bootstrap procedure was performed to test the null hypothesis that each site had the same underlying likelihood of prescribing testosterone to a patient. A simulated population was generated by randomly assigning patients to prescribed or not prescribed groups according to the observed overall proportion of testosterone. Site proportions were computed, and from these, the standard deviation among sites was computed. One thousand simulations were run. The standard deviations produced by the bootstrap procedure were then compared with the actual standard deviations from the directly observed data. The proportion of simulations in which simulated variation exceeded observed variation was used for hypothesis testing. All analyses were conducted using SAS version 9.3 (SAS Institute Inc, Cary, North Carolina).
The characteristics of the 5,196,156 male patients included in our study are reported in Table 1. The median age of this primarily non-Hispanic white sample was 63 years. Hypertension (50.8%) and dyslipidemia (46.3%) were the most prevalent physical comorbidities; 23.5% had diabetes mellitus. Approximately 15% of patients suffered from major depression.
Exogenous Testosterone Use
Overall, 85,097 men (1.6% of total sample) received exogenous testosterone from a VHA pharmacy in FY11. Table 2 presents the proportions, costs, and quantity prescribed for different routes of administration of testosterone. Testosterone injections were the most commonly prescribed form of administration, representing 61% of all prescriptions. Transdermal testosterone patches accounted for 28% of prescription fills, while the topical gels accounted for 11%. Surprisingly, a small fraction of prescriptions (0.1%) were for oral methyl testosterone, whose use is no longer recommended. Most prescriptions for testosterone formulations 79.2%) had a duration between 28 and 59 days.
Age-Adjusted Distribution of Testosterone Users
The proportion of patients on testosterone by 10-year age groups is shown in Figure 1. The distribution is peaked, with an increase from age 20 years to a peak between 50 and 69 years, followed by a decline starting at age 70 years. Among patients receiving testosterone therapy, 23% were 50 to 59 years old, 44% were 60 to 69 years old, and the majority of patients receiving testosterone (85%) were more than 50 years old. The proportion of patients more than 50 years old on testosterone was 1.7%, while the proportion patients younger than 50 years old was 1.3% (data not shown).
Figure 2 illustrates the age-adjusted percentages of patients on testosterone among the 129 VHA sites. The age-standardized proportion of males receiving testosterone by site varied from 0.3% to 3.7%, with a median of 1.4%. Our bootstrap procedure indicated that the 129 sites of care in the VHA did not have a uniform age-adjusted rate of dispensing exogenous testosterone (P <.001 for nonequivalence).
Our systematic investigation of the age-adjusted rates of exogenous testosterone use at the site of care level within the VHA, the nation’s largest integrated healthcare system, revealed several interesting findings. First, the sites of care varied in their age-standardized rates of dispensing testosterone from a low of 0.3% to a high of 3.7%, a more than 10-fold variation. Second, intramuscular injections were the most frequently used as well as the least expensive route of administration in our study, although we did not consider the expense of actually administering the injections. In the testosterone replacement therapy literature, these injections have been cited as the most widely used forms of exogenous testosterone administration due to their low cost and the convenience of infrequent dosing.21 However, the use of transdermal gels and patches is growing; it is possible that the prevalence of transdermal gels is higher outside the VHA system than it is within the VHA system. It is, however, surprising that in spite of poor efficacy and potential for liver toxicity1,2 and the recommendations of expert panels not to use oral methyl testosterone,1,2 this formulation was prescribed to more than 500 men. In our study, males between the ages of 50 and 70 years were the most frequent recipients of exogenous testosterone prescriptions. These results suggest, although they do not prove, that a proportion of patients were likely receiving testosterone therapy for age-related decline in testosterone levels, which is not an approved indication for testosterone therapy, as the risks and benefits of testosterone therapy for this group of patients have yet to be established in randomized trials.1,2,26-31
The finding that sites vary in their rates of testosterone dispensing is not surprising, because practice variation is the rule rather than the exception. However, the more interesting question relates to the underlying causes of this variation. Performance variation implies a shortcoming of the evidence base or its inconsistent application to clinical care; in this case, both may be occurring. It is reasonable to expect that better evidence and its more consistent application would lead to more uniform rates of use.32,33
Possible explanations for the observed variation in age-standardized rates of testosterone prescribing could include site-level differences (eg, region of the United States, size, academic affiliation) or patient characteristics (eg, existing physical comorbidities, race). For example, sites with high rates might have a higher number of patients with sexual dysfunction, reduced physical performance, or diabetes. Each of these physical conditions has been shown to be associated with low endogenous testosterone levels in males,11-20 and there is considerable off-label of testosterone in these conditions; therefore, variations in these factors could contribute to uneven rates of exogenous testosterone use by site. However, we would point out that the benefits of exogenous testosterone are equally uncertain for all of these patient groups. Testosterone therapy is neither indicated nor approved for these conditions, further emphasizing the importance of a more consistent policyacross the VHA system.
PDF is available on the last page.